This is a submission for the GitHub Copilot CLI Challenge
What I Built
I built FlappyHand, a hands-free interactive game where you control the character using hand gestures captured by your webcam. Inspired by the classic Flappy Bird, this game uses computer vision to track your hand state:
- Open Palm: The character flaps and flies upward.
- Closed Fist: The character descends due to gravity.
The application is built with Next.js and React, leveraging MediaPipe for real-time hand tracking directly in the browser. It features a scoring system, high score leaderboard, and smooth canvas-based rendering.
Demo
Web APP: https://sk-flappy.vercel.app/
-- simple Rule just open palm it will go up and fist go down.
My Experience with GitHub Copilot CLI
Building this project with GitHub Copilot CLI was an incredibly streamlined experience.
I really enjoyed using the GitHub Copilot CLI—it was smooth and handled tasks reliably. I started the project with only a vague prompt, and after 3–4 minor tweaks, the full app was up and running, effectively turning my idea into reality.
I used different models at different stages of the process:
- For refining the idea and creating a solid plan, I used
Gemini 3 Pro (Preview). - For development, I used
Claude Opus 4.5 - For tuning used
GPT-5.2-Codex.
Top comments (1)
CV controllers are under-explored! Cool project!